MachineSense: detecting and monitoring active machines using smart phone

Abstract

Comparison of actual running machine and recognized running machine using our prototype system for 25 minutes of time. Multivariate Gaussian Acoustic Model Naive Bayes classifier with equal prior probability Circle 1,2,3 shows the position of a microwave, a fan and a vacuum cleaner respectively at our lab office. Square 1,2,3,4 and 5 shows the position from we have identified the current running machine using our prototype application. We develop a prototype system in Android phone (Nexus S)